Moving between printers gives big difference in prints

[Herbc]

I usually print my own stuff on paper and TRY to print on canvas on my 3800, but when I need a larger print,
my local shop has been quite helpful, especially on canvas.  Their work was always directly compatible with mine.  I take a file to them, they print it on their monster Epson, no problem.
Lately, the prints from this approach are delivering VERY light prints, not acceptable.  They are quite willing to work with me, but the only solution I can think of is to take them a paper print and ask them to match it on their machine with their canvas effort; i.e. when I cannot make my machine print the canvas.
I am hoping there is a way of matching output curves or something similar so as to not have to go the trial and error route.  They use Windows machines, my work is on a Mac.
Any ideas here??

[Mark D Segal]

Windows or Mac shouldn't make a difference. It sounds like a profiling issue. Are you soft-proofing your images for their printer/paper combination?

[Herbc]

Ok, tell me what soft proofing is, and I can answer.  I do take a paper print to the shop so they can match the canvas print to it.  Obviously you won't be getting exactly the same look from say Baryata paper and glossy canvas, but the differences we are getting are huge, to the point of having to junk the canvas.  I spoke with the printer again today and he claimed he could not reproduce a canvas print today to match a canvas print he did last year from the same file due to lot differences in ink and canvas.  Considering the differences we are seeing, I can't believe this.

[Herbc]

Duh- got the tutorial on soft proofing and will study it.  My guy at the shop may (not likely) know how to do this.

Thanks

[Mark D Segal]

Time to change printers.

[digitaldog]

If you use the same media, profile and settings, the results should be visually identical. The Epson's are that close. Something is off, where? Got to check all the settings in both locations. Ideally send a target of patches to each and measure to see where there are differences and in this example, bypassing a profile and some other settings to get to the root of the output differences. That be a last resort if you find everything else in both locations is in sync.

[Mark D Segal]

Probably less time and grief expended finding a competent printer.

[TSJ1927]

This may be a time for a "New Guy" or a new shop.  If they were that light the "Guy" at the shop should have called you and told you that they appeared extremely light and there may be a problem.  Too many shops out there that bring up the file and hit "Print".   Kind of like the machine prints of old.

[Herbc]

Indeed time for a new shop.  Any recommendations?  I am in North Carolina,

[digitaldog]

Duh- got the tutorial on soft proofing and will study it.  My guy at the shop may (not likely) know how to do this.

They don't have to, you do.

[Some Guy]

I have two Epson printers of the same model and they print differently enough that I need a paper profile unique to both.  Maybe purchased two years apart, but same model.  Sort of irksome, but I live with it.

Some webinar by x-rite even mentioned that no two monitors being alike too.  Eizo might be better out of the factory in calibration around the areas of the screen, but even trying to calibrate your own with consumer-grade hardware would be tough since the factory can do maybe 9 screen areas for calibration to our one.  Example would be the edges of one screen higher Kelvin (or bluer) than edges of second, even though their centers might be similar once calibrated with our consumer stuff.

Don't even get me started with Apple and their printing ideas on their OS X system upgrades: "Our way, or no way."  Windows may be the better printer OS now, imho.

SG

[digitaldog]

When I built the Epson profiles for Epson for their EFP, we measured 5000 patches from four different printers around the country (2400, 3800, 4800, 7800, 9800 then 80 series of each). The average deltaE was less than half of one! That's tiny. So if you folks are seeing visual differences that means:

1. Newer Epson's are not behaving as older one's did (kind of doubt that, the canned profiles from Epson work pretty well).
2. There's some issue with 3rd party paper consistency?
3. There's some issue with OS and app consistancy?
4. User error.

The only way to understand what's going on, as I suggested, is printing without any color management, a set of targets, measuring and comparing.

[Mark D Segal]

Andrew,
My colour management is usually extremely robust and produces highly predictable results. As you know, I recently ran into several issues. That happened only after OSX and LR updates, and the installation of unknown updates (literally, because in Mavericks the "Updates are Ready to Install" advisory did not give me the option to unpack them). So I suspect that regardless of how well one profiles displays and printers, these application updates are perhaps doing things under the hood we (or at least some of us :-) ) don't understand. Perhaps re-profiling when outcomes break is the best corrective.

[digitaldog]

Perhaps re-profiling when outcomes break is the best corrective.

When all else fails but the only way to understand what's causing the differences is to do so scientifically. When sending targets out for a profile, you're turning off a number of areas if potential issues and if you see differences, you know where not to look. You can also see exactly where in color space the changes occur. So for example, if you see white being quite different, and we know white affects lots of colors too, you can drill down and examine what's going on. Always a good idea to keep the targets you used in the past to compare to those you print in the future. With a good running Pro Epson printer, using a properly running OS, the dE differences between units should, repeat should be tiny. But that's hardware alone. The software can and has mucked up output for various users for years.

[Mark D Segal]

Andrew, I read this over three times and I'm stymied. I don't understand what you are trying to say - honestly. I agree with you that Epson printers are made to a high standard of uniformity within a model, and that is why they can offer canned profiles that on the whole perform very well. But what you are trying to say about diagnostics doesn't provide guidance on exactly how to proceed, what one should be looking to find out, and if we can determine anything, what one can do about it short of re-profiling, because we have no control over what goes on under the hood other than not installing anything as long as what we have does what we want from it.

[digitaldog]

But what you are trying to say about diagnostics doesn't provide guidance on exactly how to proceed, what one should be looking to find out, and if we can determine anything, what one can do about it short of re-profiling, because we have no control over what goes on under the hood other than not installing anything as long as what we have does what we want from it.

IF you can reprofile, you can print and measure the target. If you can do that, you can compare the previous target measurements to the new measurements. Or you can compare the new measurement with what the older profile predicts. You have to have the tools and knowledge to do this of course.

Printer behaved one way a year ago, it doesn't today. Why? Paper, inks, driver, etc. You can bypass some of this using the above process. You can see exactly where in color space the differences show up. That aids in understanding what happened and how to fix it.

How many Epson pro printers are out there whereby users are having no issue using the canned profiles supplied by Epson? Lots. Why this works is because all those printers behave very closely (0.5 dE in my examples). If your's doesn't, you should figure out why. Comparing spectral or Lab data measurements can point you into the direction of what changed and where.

[Mark D Segal]

OK, that adds a lot of clarity; thanks. Personally, I can do those measurements. A great many people are not so equipped. Even so, having done the measurements, I'd know what changed, but I may not know exactly why, hence whether there is anything I can really do about it; if it boils down to a reversible software install issue, perhaps it can be uninstalled; otherwise the solution may simply be re-profiling for the new environment. Another option altogether is using an application that bypasses both Colorsync and the Epson driver completely - for example, LaserSoft Imaging's new PrinTao, or certain RIPs. There the technology is all self-contained and the vendor is one window of knowledge, control and responsibility.

[digitaldog]

Even so, having done the measurements, I'd know what changed, but I may not know exactly why, hence whether there is anything I can really do about it; if it boils down to a reversible software install issue, perhaps it can be uninstalled; otherwise the solution may simply be re-profiling for the new environment.

With the measurement data and something like ColorThink, you could sort by worst dE and see where in color space the big differences were showing up. That could give you idea if one or more inks is a possible issue, if the paper white is way off. If you compare the gamut, you might see areas that are clipped indicating other possible issues, media settings etc. The evaluation isn't going to tell you everything but it might help in drilling down the differences. You have to print and measure to build a new profile, so examining the new data versus the old only takes a few minutes and might indicate a fix such you don't have to reprofile. You could print the targets from another machine and compare to see if this is machine specific. You could print the targets from another application and again see if it's application specific.

[Mark D Segal]

I agree with all that Andrew - very useful guidance.

What's haunting me more than anything else when I see sudden performance change in a well-color-managed workflow is the possible impact of software changes that come from this continuous treadmill of upgrades. One is at the mercy of the attention these companies pay - or not - to respecting colour management practices and standards; as many of us seen, it is not always a seamless ride.

[digitaldog]

What's haunting me more than anything else when I see sudden performance change in a well-color-managed workflow is the possible impact of software changes that come from this continuous treadmill of upgrades. One is at the mercy of the attention these companies pay - or not - to respecting colour management practices and standards; as many of us seen, it is not always a seamless ride.

Yup, it's really frustrating and it was more common, at least for me back in the OS X 10.5/10.6 days. Apple, Adobe and Epson were all fixing other's issues and breaking things all the time. Seems better recently. It's a good excuse to have an older machine with an older OS for testing purposes.